filipe r. cogo

Latest updates

[February 2024] Our paper entitled "Towards Semantic Versioning of Open Pre-Trained Language Model Releases on Hugging Face" was accepted for publication on EMSE! This paper identifies a lack of versioning good practices regarding the releases of pre-trained language models in the HuggingFace ecosystem and discusses potential solutions to the identified problems.

[January 2024] Our paper entitled "An Empirical Study on Vulnerability Disclosure Management of Open Source Software Systems" was accepted for publication on TOSEM! Based on a survey with open-source contributors, we describe the adopted vulnerability disclosure practices and the perceived advantages and disadvantages.

[January 2024] Our paper entitled "On the Workflows and Smells of Leaderboard Operations (LBOps): An Exploratory Study of Foundation Model Leaderboards" was accepted for publication on TSE! The paper focuses on understanding how FM leaderboards operate in real-world scenarios and identifying potential pitfalls and areas for improvement.

[January 2024] Happy to share that our paper entitled "InterTrans: Leveraging Transitive Intermediate Translations to Enhance LLM-based Code Translation" was accepted on ICSE'25! In this paper, we introduce the Tree of Code Translation (ToCT) technique that leverages the multilingual capabilities of LLMs and achieves state-of-the-art (SOTA) performance across CodeNet, HumanEval-X, and TransCoder datasets for code translation.

[January 2024] We will have a budget-friendly, mini version of the AIware Leadership Bootcamp co-located with ICSE 2025 at the University of Ottawa on May 3-4, 2025. More info here.

[November 2024] I am lecturing a prompt engineering session at the super comprehensive AIware Leadership Bootcamp happening in Toronto. All the teaching materials can be found at the event's website.

[November 2024] Together with my colleague Dayi Lin, I will present a tutorial on Software Engineering for FMware at the ASE'24 conference in Sacramento, USA.

[September 2024] I am co-organizing the AIware Latam event with my colleagues César França, Cleidson de Souza, Igor Scaliante Wiese, and Igor Steinmacher. This is part of a series of events that reunite industry and academia to discuss the challenges and opportunities at the intersection between software engineering and foundation models. The event has also successfully established an AIware steering committee and the organization committee of AIware Latam'25.

[July 2024] I'm happy to share that our paper entitled "Mitigating the Uncertainty and Imprecision of Log-Based Code Coverage Without Requiring Additional Logging Statements" was accepted for publication at TSE! This paper is the result of Sean's master's thesis and a collaboration between the Centre for Software Excellence and the University of Waterloo.

[July 2024] I am helping co-organizing the 1st ACM International Conference on AI-powered Software (AIware) as a Local Chair. The event is co-located with FSE in the beautiful city of Porto de Galinhas, Brazil.

[April 2024] Our paper, "Rethinking Software Engineering in the Era of Foundation Models," was accepted into the FSE Industry Paper track! In it, we reflect on 10 software engineering challenges to developing foundation models and their applications.

[April 2024] I'm honoured that our paper entitled "Exploring the Impact of the Output Format on the Evaluation of Large Language Models for Code Translation" received the best paper award at FORGE'24.

[March 2024] My colleagues and I will give a tutorial entitled "Software Engineering for AIware" at FSE 2024. In addition to presenting the latest research and industrial practices in engineering foundation models and their applications, we will have a hands-on session to demonstrate the deployment and use of foundation models in consumer computing infrastructure!

[February 2024] Our paper entitled "Exploring the Impact of the Output Format on the Evaluation of Large Language Models for Code Translation" was accepted at FORGE'24. Our paper demonstrates that controlling output format is paramount to avoid bias during the evaluation of LLMs for code translation.

[January 2024] My colleagues and I gave a technical brief entitled "Software Engineering for AIware" at ICSE 2024. We presented some of the latest research and industrial practices in engineering foundation models and their applications.

[January 2024] Our paper, "Studying the Impact of Risk Assessment Analytics on Risk Awareness and Code Review Performance," was accepted for publication with EMSE! This paper uses a controlled experiment to evaluate Gherald, a risk awareness prototype plugin for code review interfaces.

[November 2023] I participated in the "FM+SE Vision 2030" workshop in Mexico City. It was great to hear the discussions about the future of foundation models and software engineering from top scholars and practitioners worldwide!

[September 2023] I am excited to teach the "Software Engineering and Foundation Models" course in Fall 2023 at the Department of Computing, Queen's University.